Knowledge tracking method and system

Abstract

The invention relates to a knowledge tracking method and system, and belongs to the technical field of knowledge tracking. Comprising the following steps: constructing a DMKT model based on a DKT model; the DMKT model comprises an input layer which is used for obtaining a coding vector according to student answer data and domain feature coding; a hidden layer which is used for obtaining a hidden layer output result according to the coding vector, the knowledge state data of the students at the previous moment and the domain feature codes; an output layer which is used for obtaining a prediction result according to an output result of the hidden layer; and a multi-knowledge-point mapping layer which is used for obtaining a multi-knowledge-point mapping result according to the prediction result; obtaining historical student answer data, historical domain feature codes and a historical prediction result, and training a DMKT model in combination with a multi-knowledge-point mapping result; and outputting a prediction result at the next moment according to the trained DMKT model. The problems of lack of domain feature fusion and feature reduction in the fusion process are solved, meanwhile, the occurrence of knowledge point association relation forgetting is inhibited, and accurate tracking of the knowledge level of students is realized.

Description

technical field [0001] The present invention relates to the technical field of knowledge tracking, in particular to a knowledge tracking method and system. Background technique [0002] In recent years, the wide application of online learning platforms and intelligent tutoring systems has provided students with a wealth of practice test questions, one of which may be related to one or more knowledge points. The probability of a student correctly answering a test question depends on his state of knowledge, that is, the degree of mastery of knowledge points. The purpose of knowledge tracking is to track students' knowledge status, and predict the probability of students answering the next test question correctly based on the students' answer records. However, there are existing knowledge tracking methods such as Deep Knowledge Tracing (DKT), whose input only includes question labels and correct answers, and the answer results are obviously affected by other domain characteris...

AI Technical Summary

Problems solved by technology

However, there are existing knowledge tracking methods such as Deep Knowledge Tracing (DKT), whose input only includes question labels and correct answers, and the answer results are obviously affected by other domain characteristics such as the number of answers, answering time, etc.

Although some methods have tried to integrate features into the DKT model, there is a problem of low prediction accuracy. The main reason is that the key information reduction problem of features in network transmission is not considered.

[0003] In addition, the cyclic neural network in knowledge tracking itself has the problem of long-term dependency forgetting, that is, the network will forget what it has learned before, which will lead to the loss of some key information; and there are a large number of multi-knowledge point problems in knowledge tracking, and there are also problems between problems. There are complex knowledge point association relationships, and the forgetting of these relationships will cause the network to fit wrong features, resulting in wrong associations and continuous offsets of knowledge points

Images